Blind Research??

What is blind research? I'm assuming it's your scientists not knowing what they're discovering until it actually has been discovered.
 
Correct, its more realistic that way. It must be possible to get the ai to automatically pick the next tech much the same way that the ai picks its own tech choices.
 
Not really, because you are unable to make a beeline for ironworking for example, you have to work with what you have!

Ideally it would be cool if you can specify what the priorities should be "eg focus on military" or focus on production.

I wanted to use it on my smac mod that we were workign on....
 
The negative of this idea is that competitive game play won't be fair since the tech choices will be random. Someone might accidentally beeline to a great unit, while someone trying to warmonger might keep rolling economic techs.
 
GoodGame said:
The negative of this idea is that competitive game play won't be fair since the tech choices will be random. Someone might accidentally beeline to a great unit, while someone trying to warmonger might keep rolling economic techs.
Click to expand...

Well, wait a second... now, presumably, you wouldn't have it so that your technologies are completely random... rather you would invest research in a particular field. So the warmonger would focus on a military field of research and be rewarded with military techs albeit maybe not the specific military tech they want.

But also, I would recommend making it so that there is no longer just an abstract research amount... different buildings, units, etc. should increase research in particular areas. For example, an Observatory wouldn't give a blanket boost in research, it would only increase research in say the Science field or whatever it would be called.. whereas Marketplaces might increase research in Economics and Barracks in Military.

So, yes, it takes away the ability of a player to hand-pick the rewards from their research efforts, but it maintains choice in that you determine where you're going to focus your research efforts and can actually increase research output for your focused category depending on the kind of society you build via techs, civics and units. In fact, if anything, it gives additional choices to be made since focusing purely on the military will mean sacrificing the economic and thus slowing the process of developing economic techs, which presumably you will HAVE to do at some point unless you are able to strictly get those techs through intimidation and conquest.
 
Just to give an idea for the algorithm, you don't need to necessarily have it be random to have it be blind. For example, you could take the costs given and multiply that cost (just for this algorithm, that is) based on priorities - a "Militaristic" research priority, for example, might multiply by 0.5 for a military tech, but a peaceful priority might multiply by 2.0 for the same tech. Then research the "cheapest" tech.
All other things being equal, the tech sample at any time should be fairly even across the board - everyone should be in the same general era, but with militaristic civs being higher up on militaristic techs, economic civs with econmic techs, etc. No-one will have high tech computers with no knowledge of agriculture, though.
 
Victarus, that would be a great algorithm, however do you know that the system has already got a built in ai weighting for each tech? the priority can simply double the weighting.

Where in the code is the python for allowing player tech choices?
 
troyDoogle7 said:
Victarus, that would be a great algorithm, however do you know that the system has already got a built in ai weighting for each tech? the priority can simply double the weighting.
Click to expand...
Then why the debate? Don't give me the idea that ideas are needed and then go "Oh, we're done with that part". :D
As for the algorithm, it's based on getting the "best" technology. I was thinking of it being more even across the board since there wouldn't be that competition to get the best tech - since no-one's picking technologies, there's no need to use a complicated (and just a little sketchy...) algorithm to do the job.


As for Python, you're on your own there. I'm a C++ guy looking at making a mod, not the other way around - I still haven't managed to look through the majority of the SDK yet, let alone get a firm grasp on it and move on to the next section.
 
If that's the case then it's playable (basically it's semi-blind research), but definitely some purists will dislike this since some of the 'chess' of civ is even in the tech research order.

Dom Pedro II said:
Well, wait a second... now, presumably, you wouldn't have it so that your technologies are completely random... rather you would invest research in a particular field. So the warmonger would focus on a military field of research and be rewarded with military techs albeit maybe not the specific military tech they want.
Click to expand...
 
GoodGame said:
If that's the case then it's playable (basically it's semi-blind research), but definitely some purists will dislike this since some of the 'chess' of civ is even in the tech research order.
Click to expand...

And that is why there is a vanilla Civ4 ;)
 
I've long been dissatisfied with the research system in Civ. Don't know if blind research is the way to go, but I do like the idea of breaking things up into categories--more so than is currently done.

I suppose you could have different flavors for each tech, e.g. currency is a strong (10 on a scale from 1 to 10) flavor in economics, but a weak (1) in military. That way, if you invest in specific areas, you're more likely to discover a tech that is weighted toward that area. Is this what you were basically saying already?
 
Gaius Octavius said:
I suppose you could have different flavors for each tech, e.g. currency is a strong (10 on a scale from 1 to 10) flavor in economics, but a weak (1) in military. That way, if you invest in specific areas, you're more likely to discover a tech that is weighted toward that area. Is this what you were basically saying already?
Click to expand...

I don't even know what I'm saying... I'd like to have it a little more directed than that in that when you discover a new tech and it has the pop up asking what you'd like to research next, you'd pick a category rather than a specific tech... however, I would like it if the direction of your economy might benefit research time for those categories... to be honest, I haven't really given it a tremendous amount of thought. I've never thought the current system bad enough to make it a priority... but I'm all in favor for trying something new.
 
CvPlayerAI.cpp, lines 1830-2690:
CvPlayerAI::AI_bestTech()
CvPlayerAI::AI_chooseFreeTech()
CvPlayerAI::AI_chooseResearch()

AI_bestTech() is probably the one you're looking for.
 
Actually, it can be done in python. You just need a double tech tree.

First, disable all default techs, and preferably hide them. Then create new techs, called something like Military A1, Economics A1, etc, like in Alpha Centauri. These techs don't give anything, they are just flavoured so the AI might research what it needs.

Now, once TECH_MIL1 is researched, tell python to randomly pick a pre-defined tech & give it to that player. It should also check if the tech it picked isn't already known by that civ.

If you need the coding for this, I think I might find the time to do so, but I give no assurance of that happening.
 
You must log in or register to reply here.

Similar threads

N
Merging a Mod with a Map
Replies
5
Views
498
Nortabano
N
O
Tool - Civ4 Map-Editor V3 - third test release
Replies
15
Views
1K
herve.moravio_aposto
H
O
Tool - Civ4 Map Cleaner
Replies
0
Views
138
onkelm
O
Perdedor
Link For Custom Leader Heads Is Broken
Replies
0
Views
355
Perdedor
Perdedor
J
Mod for slowing research rate?
Replies
2
Views
849
jerminor
J
Back
Top Bottom